The genetic etiology of Autism Spectrum Disorders (ASDs) reflects of the nature of genome instability. Based in part on studies of copy number variation (CNV) from the first phase of this project, it is now widely recognized that de novo or recent mutations in the germline are important contributors to risk for ASDs. Mutations that confer high risk for autism are located within regions of hypermutability (ROHs). Examples include large >500 kb regions at 1q21.1, 15q11-13 and 16p11.2, where local genomic architecture predisposes to frequent recurrent rearrangements. All told, de novo CNVs contribute in approximately 5-10% of cases. It is likely that other classes of de novo mutation (DNM), in exons and in non-coding regions, contribute to ASD. As a field, we have a limited knowledge of global patterns of germline mutation in humans, and we have an incomplete understanding of how DNM contributes to risk for disease. Based on the knowledge above, and based on key findings that have emerged from our preliminary genome sequencing studies in autism, we hypothesize that, regions of hypermutability (ROHs), primarily consisting of nucleotide substitution hotspots, play a major role in ASDs. In the new phase of our project, we will apply a whole genome sequencing-based approach to determine the role of ROHs in ASD. We will investigate the nature intrinsic hypermutability and the extrinsic forces, such as paternal age, that influence rates of germline mutation. We will accomplish these goals through the following specific aims: Specific Aim 1 will characterize germline de novo mutations (DNMs) by whole genome sequencing in families. These studies will identify and validate ~8,000 de novo point mutations and structural variants in a sample of 120 trios (60 ASD and 60 controls) and determine the parent of origin of DNMs. Specific Aim 2 will investigate the role of ROHs in ASD. We will identify ROHs based on the regional density of DNMs in the genome, and determine the effects of DNA sequence features on rates of mutation. We will determine the association of ROHs with ASD in the discovery sample and in exome and CNV datasets from an independent sample of 2700 cases and 2700 controls. Specific Aim 3 will characterize the effects of extrinsic factors, including parental age and environment, on genome-wide rates of mutation. We will quantify the effect of paternal age on pathogenic and neutral alleles in sperm and investigate whether some DNMs confer a germline selective advantage. The findings of this study will provide fundamental insights into the genetic basis of autism risk and the genetic mechanism of the observed parental age effects in ASD. We will identify genes that confer significant risk for autism, and we will determine how intrinsic properties of the genome interact with extrinsic forces to determining risk for disease in offspring.